ELECTRICAL LOAD DEMAND MODEL BY NEURAL NETWORK METHOD

摘要

Usage of neural network for estimation electrical load demand has been started for years. In this method, Time series variable is analyzed for projection of future. In this paper, we will estimate electrical load demand for long-term and determinate effective parameters. Effective parameter already has been shown in previous studies of author. This paper, First by using back propagation method and considering of effective input for electrical load demand, such as income, value added of Industry and home sector, price of electricity and other carrier energy, population, historical and economical effects, primary neural network model with input, hidden and output layer will be designed. Then weights will be given in learning process. Finally it leads to optimized network for estimation of electrical load demand. The Final model is useful for policy maker for five-year programs in Iran.